Comparison of Diets for Mass-RearingAcheta dornesticzs (Orthoptera: Gryllidae)

as a Novelty Foodo and Comparison ofFood Conversibri Efficiency with

Values Reported for Livestock

BARBARA J. NAKAGAKI ENO GENE R. DBFOLIART

Department of Entomology, University of Wisconsin, Madison, Wisconsin 53706

J. Econ. Entomol. 84(3): 89f-896 (f99f)ABSTRACT As part of research on mass-rearing the cricket Acheta domesticus (L.) as anovelty (innovative) food, four cricket diets, two prepared in the laboratory and two com-mercial, were compared on the basis of cost per kilogram (wet weight) of eighth instarsproduced. Costs were influenced by dietary ingredients, mean cricket wet weight at time ofharvest, and feed/gain tatios. For the laboratory-prepared diets, crickets grown on Patton'sdiet no. 16 or on NRC reference chick diet averaged 0.443 and 0.418 g at time of harvest,with feed/gain ratios of 0.923 and 0.949, respectively. Because of the cost of ingredients,however, the cost per kilogram of live crickets produced was only $0.21 for the NRC chickdiet compared with $2.55 for Patton's no. 16; these costs exclude the cost of labor for mixingthe diets. Perfoimance of crickets on the laboratory-prepared diets was somewhat betterthan performance on the commercial diets. The food conversion efficiency of crickets atSffC or higher was found to be higher than reported for broiler chicks and pigs and muchhigher than those reported for sheep and cattle.

KEY WORDS Insecta, Acheta domesticus, human food, diets

Snvnner- spEcIES op crickets are used as food bynon-European cultures, notable among which areBrachytrupes portentosus (Lichtenstein) through-out southern Asia and BrachEtrupes membrana-ceus Drury throughout eastern Africa (DeFoliart1989). Although crickets are not usually thought ofas food for humans in the United States, the housecricket, Acheta domesticus (L.), is one of severalinsects recommended by Taylor & Carter (1976)for inclusion in their gourmet recipes. Nutrition-ally, the cricket has been shown to be a proteinsource of high quality in feeding trials with chicks(Nakagaki et al. 1987) and weanling rats (Finke etal. 1989). As a result of these attributes, we areinvestigating the mass production of A. domesticusas a novelty food (a new, unusual, or innovativefood) for human use.

Our main objective in the work reported herewas to determine which of several available cricketfeeds could be used to produce crickets of har-vestable age at the lowest possible cost. Ifi our ear-lier rearing of crickets for use in animal feedingtrials, we used Patton's diet no. 16 as the feed forcrickets. Of 16 oligidic diets tested by Patton (1967),this diet produced the best overall cricket perfor-mance. The ingredients used to prepare the diet,however, make it quite expensive ($2.76 per kg offeed) for use in mass rearing crickets for experi-mental animal feeding trials and far too expensivefor use on a commercial scale. Preliminary cricketfeeding trials with a formulation prepared accord-

ing to specifications of the National Research Coun-cil (NRC) reference chick diet (National ResearchCouncil 1977) gave good results and led to thecricket-feeding experiments reported here. In ad-dition to Patton's diet no. 16 and the NRC chickdiet, a commercial cricket feed and a commercialrabbit chow were included in the tests.

The results of our tests permit a comparison offeed costs in cricket production and also a com-parison of the food conversion efficiency (ECI) ofthe cricket-with those reported for conventionalvertebrate livestock. Wet-weight ECIs (ECI :[weight gained/weight food ingested] x 100[Waldbauer 1968]) were recorded for two reasons.First, crickets are shipped alive to buyers, and rec-ipes (Taylor & Carter 1976) are based on the useof freshly killed or frozen crickets. Second, feed/gain ratios of livestock are determined on a wholebody, live-weight basis; thus cricket wet weightswere needed for direct comparison of crickets andlivestock sent to.market, We have also attemptedcomparisons of both wet and dry weight ECIs whenlosses due to dressing percentage and carcass trimare included. '

Materials and Methods

In all experiments except one, cricket feedersand housing materials were made of plastic. Wa-terers were glass vials filled with deionized waterand plugged with dental rolls. Except in the ex-

O022-O4SS / 97 / 089f -0896$02.00/0 O l99r Entomological Society of America

892 JounNal or EcoNouIc ENrouol-ocv Vol. 84, no.3

Ta}le l. Proximate analyses, calculated netabolizable energyr and cost of experimenlal diets fed to crickets

% Crudeprotein

Zo Crude fat % Ash% Crud. Metabolizable

fiber i:"r/t1;Cost,s/kg

Patton's no. 16NRC reference chick dietSelph's cricket feedPurina Rabbit Chow

30.5

17.0p14.0h

' 5.2J,D3.5,2.& t0.0e

2.84.94.U

20.v

3,r563,rI83,0002,180

2.7tr0.224Lzld0.38e

5.r4.4

d Cost of ingredients only; cost of labor for mixing diets not included

' Label-guaranted minimum.c Label-guaranteed maximum.d Purchased in 4.55-kg lots.

" Purchared in 22.73-kg lots.

periment comparing performance on different di-ets, all crickets were fed the NRC reference chickdiet described below. AII diets were ground in aWiley mill to a 20-mesh particle size, and all ex-periments were conducted under a 24 h daylightregimen, At the conclusion of experiments, cricketswere killed by freezing and then weighed.

To harvest the maximum proportion of large,unwinged nymphs, it was necessary to determinethe rate at which adult crickets appear in a cohortat a given temperature. An experiment was con-ducted in l9-liter aquaria at 35 t 2oC, using livereplicates of 200 crickets each fed the NRC stan-dard reference chick diet for 40 d. The proportionreaching the adult stage was tabulated at 3-d in-tervals following the fiist appearance df adults.

To determine ECIs of cricket cohorts fed thedifferent experimental diets, newly hatched crick-ets were fed Patton's modified diet no. 16 for 3 dbefore transfer to vented plastic cages (8 by 18 by4 cm). The experiment was conducted at 33 + 2'C.Fifteen replicates of l0 crickets per cage were fedone of four diets for 2I d; crickets were harvestedat 24 d of age (for proximate analyses and calcu-lated metabolizable energy of the diets, see Tablef ). The diets are as follows:

Patton's modifted diet no. 16. Contains 30 g soy-bean meal (4I% protein), 25 g wheat middlings,15 g dried skim milk, 10 g corn meal, l0 g driedbrewer's yeast, and 10 g de{atted liver powder.This diet was slightly modified in our test by theaddition of 0.5 g cholesterol.

Purina Rabbit Chow (Purina Mills, St. Louis,Mo.). A pelleted commercially available mainte-nance diet for nonbreeding rabbits.

Selph's Cricket Feed (Selph's Cricket Ranch, Inc.,Memphis, Tenn.). A commercially hvailable brick-et diet.

NRC reference chick diet. Ingredients are pres-ent in the following percentages: Ground yellowcorn, 58.0; soybean meal, 35.0; corn oil, 3.0; cal-cium carbonate, 1.0; dicalcium phosphate, 2.0; io-dized salt, 0.5; Dl-methionine, 0.25; and cholinechloride (50Vo), O.15. The following ingredients aremC/kC diet: Manganese sulfate ' 5Hr0, 170.0; zincsulfate ' HrO, 110.0; ferric citrate ' 5HrO, 500.0;copper sulfate ' 5HrO, 16.0; sodium selenite, 0.2;

thiamine-HCl, I.8; riboflavin, 3.6; calcium panto-thenate, 10.0; niacin, 25; pyridoxine-HCl, 3.0; fo-lacin, 0.55; biotin, 0.15; vitamin B,r, 0.01; and vi-tamin K-1, 0.55. Vitamin A was added to the dietat a rate of 1,500 units/kg of diet, vitamin D. at400 units/kg of diet, and vitamin E at 10 units/kgof diet. The 1,2 dihydro-6-ethoxy-2,2,4-trime-thylquinoline (ethoxyquin) was omitted from thediet because Patton (1963) found that feeding med-icated diets adversely affected cricket growth.

As it would be desirable to clear the gut in crick-ets intended for human consumption, wet weightwas determined for crickets deprived of feed thelast 24 h before harvest compared with crickets notso deprived. Both groups had ad lib access to waterup to the moment of harvest. The dry weights ofcrickets in this experiment were also obtained. Eachtest group consisted of l0 replicates of 14 juvenilecrickets (7 males, 7 females). The crickets werekilled by Ireezing 4 d after the appearance of thelirst adult. The crickets were weighed, dried in a60"C drying oven, then reweighed until there wereno further changes in weight. Cricket Iegs are notusually ingested when nymphs are used as partysnacks or in recipes; therefore, the proportion ofwhole-body weight represented by the Iegs wasrecorded in this experiment.

Differences between means were determined us-ing Tukey's multiple range test for unequal samplesizes and the Student-Newman-Keuls revised LSDtest (USDA 1977).

Results and Discussion

Our system for producing crickets as a potentialnovelty food is designed for harvest of eighth (final)instars to obtain crickets before they attain theirwings. At 33-35 + 2"C, we have ordinarily har-vested crickets at 28-30 d after the eggs hatch, atwhich time only a few have reached the adult stage.In the experiments conducted to gain more precisedata on the temporal occurrence of adult cricketswithin a cohort, the mean interval to appearanceof the first adults was 26.8 + 0.44 d (i + SE)(SD,f.4). At 3 d after appearance of the first adults (orday 30 after eggs hatch), adults comprised 3.5 t0.l9Vo ol the cohort; at 33 d, 80,5 + LI6%; at 36

June 1991 NAKAGAKT & DnFor-Ianr: DIETS FoR HousE Cnrcrnr

Table 2. Veights (t + SE) and feed/gain ratios of ;i, domesticus after 2l d on experimental diets (wet-weighr basis)

893

Diet Wt crickets, g t SE Feed/gainProtein gained/

protein fed

Feed cost/kgof crickets

produced, g

Patton's no. 16NRC reference chlck dietPurina Rabbit ChowSelph's cricket feed

0.443 + 0.009a0.418 + 0.007ab0.407 + 0.007b0.406 + 0.010b

0.923 + 0.012a0.949 + 0.026a1.651 + 0.030b1.081 + 0.012a

0.745 + 0.009d0.999 + 0.026b1.139 + 0.013a0.908 + 0.015c

2.55o.2t0.63r.3l

Means within the sme column followed by different letters differ signiffcantly (P < 0.05) (see text).

d,53.1 + 3.04%; and at 39 d,79.3 + 3.53%. Onthe basis of these data (and past experience), crick-ets used in the feeding trials with experlmentaldiets were killed at 24 d after eggs hatched topreclude the appearance of any adults.

In the feeding trials with experimental diets, themodified Patton's diet no. 16, not unexpectedly,produced the largest crickets (mean wet weightOAB g) and the best feed: gain ratio (0.923) (Table2). The diet is very high in protein and containsliver powder which is thought to contain a growthfactor (Neville et al. 196I, Patton 1967). The size(F : 4.29; treatment df : 3, error df : 56; P <0.05) and feed: gain ratio (F : 253.8; treatment df:3, error df:56; P < 0.05) of crickets fed Patton'sdiet were not signiftcantly better, however, thanthose obtained with the NRC reference chick diet(mean wet weight 0.418 g; feed: gain ratio 0.949),and the feed: gain ratio on Pattoi's was not sig-nificantly better than that obtained on Selph'sCricket Feed (I.081). Considering the low protein,low metabolizable energy, and high fiber contentof Purina Rabbit Chow, this diet performed re-markably well (Table 2). Because these nymphswere harvested at 24 d of age, they were slightlysmaller than nymphs would be at normal harvestage. In our experience, the mean weight of cricketcohorts fed NRC diet to age 28-30 d has routinelybeen about 0.45 g.

In protein efftciency (based on label informationfor the commercial products) Purina Rabbit Chowranked as best followed by the NRC diet, the Selph'sCricket Feed and Patton's no. 16 (Table 2). Alldifferences were significant (F : 21.6; treatrnentdf : 3, error df :56; P < 0.05).

From the standpoint of feed costs per kilogramof live crickets produced, the NRC reference chickdiet was clearly superior to Patton's no. 16, costingonly $0.21lkg (wet weight) compared with $2.55lkg for the latter (Table 2). These costs dre for thediet ingredients only and do not include labor;therefore, they cannot be compared directly withthe prices of the commercial diets. With the equip-ment available to us, it requires about 2 h to weigh,grind, and mix l0 kg of diet. About half of thistime is spent in grinding the materials in a Wileymill. Thus, the labor increment averages about 12min/kg of feed prepared.

Of the two commercial diets, from the stand-point of feed costs per kilogram of live crickets

produced, the Purina Rabbit Chow cost onlv $0.63/kg (wet weight) compared with $1.31/kg for Selph'sCricket Feed (Table 2). This apparent cost advan-tage is reduced somewhat, however, by the factthat Selph's is ready to feed as purchased, whereasPurina Rabbit Chow is formulated as pellets whichmust be ground to 20-mesh particle size before use.In comparatively large-scale cricket production,until there is a change in commercial availabilityof formulations, the decision whether to use hand-mixed or commercially available feed would de-pend on the individual situation relative to labor.

For commercially produced crickets sold live forffsh bait, caged bird feed, or reptile feed, the cur-rent prices are about $15/1000 with suggested re-tail prices of >$3/100. Lt 2,000-2,200 crickets/kg, the prices are about $30/kg and $60/kg forcrickets sold at wholesale and retail, respectively.From our feeding trial data, it is evident that thecost of feed ingredients represents a very smallfraction of the current price of crickets sold com-mercially. If crickets were to become popular as aparty snack, the profit potential for locally pro-duced crickets as a small-farm specialty crop is

obvious.The whole-body composition (wet weight prox-

imate analysis) of late eighth instars as found byWoodring et al. (1977) shows nymphs to be sub-stantially higher in water content, generally similarin protein level, and substantially lower in fat thanis the case in vertebrate livestock (whole body, butingesta-free) at market age (Table 3). Although thecrude protein level of fresh crickets is similar tothat of vertebrate livestock, the quality, based onNPU (net protein utilization), PER (protein efft-ciency ratio), and chemical score in relation to ami-no acid content in egg protein is lower than thatof protein from vertebrate animals and other high-quality protein sources such as whole hen's egg andcasein (e.g., Dreyer & Wehmeyer 1982, Ozimeket al. 1985). Insett crude protein is of relativelylow digestibility compared with these sources, ap-parently becatse some of the nitrogenous compo-nents are in the form of chitin in the integument,and chitinase is absent from monogastric animals.When the chitin is removed the protein qualityimproves. Ozimek et al. (1985) reported that re-moval of chitin following alkali extraction im-proved the true protein digestibility, PER, and NPUfrom 71.5%, I.50, and 42.5, respectively, in whole

894 ]ounNal or EcoNovrc ENronrolocv Vol. 84, no. 3

Table 3. Body composition of animals at market weight (vertebrates on an ingesta-free basis) and ECI of livesrockand crickets in terms of live weight gain, dressing percentaget and carcass refuse

Species andreferences

% Water % Protein Vo Fat % Wet wt % CarcassECI Dressinf refuseb

Cricket nymphWoodring et al. 1977This paper

Bioiler chickEnsminger 1980 (market)Meyer & Nelson 1963 (age 8 wk)cLovell 1979

PigEnsminger 1980Meyer & Nelson 1963cWhittemore & Elsley 1976Lovell 1979

Sheep

Ensninger 1980 (choice)Meyer & Nelson 1963Berg & Butterffeld 1976Lovell 1979

68.474.2

64.0

53.5

52.O

t7.o18.2tlt

r0.3

t4.2

!34.438.029.5

26.021.126.9

6872

67

*

oo

;48

28

3'

Y18.8,y

13.012.7,!

I00 ;

t

;

;a16 60

;;;

50.0

52.6

Ensminger 1980 (fat) 53.2 15.0 29.0Meyer & Nelson 1963c 15.2 25.2

Steer

o The marketable percentage of the animal after slaughter.,In poultry, bones only; in pork and beef, bones, trim fat, and tendons; in crickets, \egs (77% of wet weight), crude fiber (<3.020;

although not removed, fiber is here considered as refuse because of its indigestibility).c Average for animals fed either a high-fiber or low-fiber diet.

dried honey bees (Apis n'Lellif era L.) b 64.3%, 2.47,and 62, respectively, in the protein concentrate.The corresponding values for casein werc 96.8%,2.50, and 70, respectively. In addition, whole groundinsects have been shown to be an excellent sourceof protein compared with most plant sources (e.g.,Finke et al. f989).

Our data and those of Meyer & Nelson (1963)can be used to compare the wet weight ECI of A.domesticus and young vertebrate livestock fed asimilar dietary protein level (Table 3). Meyer &Nelson conducted feeding trials on both high- andlow-fiber diets containing, respectively, 21.4 and19.4% crude protein, 4.2Vo ether extract, 13.6 and'7.6V0 crtde fiber,3.7 andLT% lignin,7.0 and6.3%ash, and 4.87 and 4.85 kcal of gross energy/g (theECIs in Table 3 are an average for these two diets).Ages of the animals at the beginning of the feedingtrials and durations of the trials were: chicks I d,fed for 56 d; pigs 12 wk, fed for 90 d; sheep 5 mo,fed for 60 d, and; steers 8 mo, fed for I34 d. Asshown in Table 3, the wet-weight'Ecl of A. do-nlesticus in our trials (based on the average of thefour diets in Table 2) was far higher than the ef-ficiency found for any of the vertebrate species byMeyer & Nelson; i.e., more than twice as high asin chicks, *3 times higher than in pigs, 5 timeshigher than in sheep, and nearly 6 times higherthan in cattle. Cricket efficiency in our trials wasalso much higher than efficiencies summarized froma number of sources for terrestrial livestock byLovell (1979) (Table 3).

Meyer & Nelson chose their rations to study com-parative utilization of the same proportion of feed-stuffs by the different species and recognized thatthe rations were not necessarily the most suitablefor each species. In their tests, the pigs, sheep, andsteers were all at market weight when the trialswere terminated, whereas the chicks, although fedfor 8 wk, weighed only about half as much as themarket-ready broilers listed by Ensminger (1980)(Table 3), suggesting that neither of the diets fedby Meyer & Nelson were conducive to high foodconversion efficiency in chicks. The efficiency ofcrickets, however, was still nearly twice as high as

the much higher efficiency reported for broilers byLovell (1979).

In comparing food conversion efficiency ofcrickets rvith vertebrate livestock, two additionalfactors must be considered to obtain a true con-version efficiency from animal feed to human food.These are dressing percentage and carcass refuse(Table 3). For each class of vertebrate livestock,these two factors together reduce the edible portionof the dressed carcabs to =50% of whole-bodyweight. A comparable, but reduced, loss occurs inthe nymphal erictet carcass when legs are dis-carded and because the chitinous integument, al-though not usually removed, is indigestible by sim-ple-stomached animals. The crude fiber(integument) of the live cricket comprises <3% otits body weight, however, based on a dry weightcrude fiber content of 7.0% reported by Nakagakiet al. (1987), and the legs comprise *L7% ol total

june 1991

Table 4, Dry matter percentage of eighth instars and"carcass loss" represented by removal of legs

Crickets starved Crickets notlast 24 h starved

x

Whole BodyWet wt, gDry wt, g% Dry matter

Legs onlyWet wt, gAs % of whole bodyDry wt, gAs % of whole body

0.456 0.0060.116 0.002

25.4 0.164

0.079 0.00117.3 0.170.021 0.004

18.2 0.34

0.463 0.0080.121 0.003

26.2 0.2t

0.077 0.00r16.6 0.240.020 0.004

16.8 0.28

body weight (Table 4). In Table 3, we have assignedboth crude fiber and legs to the carcass refuse cat-egory as legs would be present'in crickets marketedalive. If marketed fresh-frozen or freeze-dried, theymight or might not have the legs removed.

The whole-body ECI margin in favor of cricketsis reduced when calculated on a dry weight ba5isbecause of the higher water content of crickets(Table 5D). The calculations in Table 5 are basedon the averages of the values given in Table 3 forwhole-body live weight, whole-body water con-tent, dressing percentage, and carcass refuse; thus,they represent rough approximations only. The dryweight of cricket nymphs was foun'd to be x26%of fresh weight in our test, =25.4Vo in cricketsdenied access to feed the last 24 h before harvestand =262% in crickets not deprived of access (Ta-ble 4). These dry-weight ratios are somewhat lowerthan the 3L6Vo dry weight found by Woodring etal. (1977) on late eighth instars, suggesting thatmany of our crickets were in the early eighth instarat time of harvest. Their data were averages forthe final five days, days 4-8, of the eighth instar;Woodring et al. (1977, 1979) have shown that watercontent in the eighth instar nymph declines froma first-day peak of 74-75.9Vo to a steady 68% by

895

the third or fourth day after ecdysis. Food con-sumption virtually ceases during the last 2-3 d ofboth the seventh and eighth instars, but water con-sumption continues and there is no loss in either

" wet or dry weight (Woodring et al. 1977, Roe etal. 1980). Woodring et al. (1979) reported an ECI,at 30'C, of 29% for the eighth instar female A.domesticus on a dry-weight basis. This ECI ob-tained on the ftnal instar can probably be consid-ered as conservatively representative for the entirenymphal period, as a high proportion of growth inmost insects takes place in the final two instars andECIs tend to be as high or higher in the earlierinstars (Waldbauer 1968).

When losses due to dressing percentage and car-cass refuse are accounted for, the adjusted dryweight ECI of crickets is still more than twice ashigh as those of broiler chicks and pigs, >4 timeshigher than sheep, and nearly 6 times higher thansteers (Table 5, E). One negative aspect is that thehigh cricket ECIs discussed here are observed onlyat temperatures of about 30oC or higher. Roe et al.(1980) found (last-instar female) dry-weight ECIsof 30 and 32 at temperatures of 30 and 35oC, re-spectively, but an ECI of only 20 at a temperatureof 25"C.

Finally, in considering feed utilization efficiencyin the broad sense, crickets have high fecundity,especially in comparison with that of ruminant her-bivores and pigs. Female A. domesticus lay anaverage of 1,200-I,500 eggs over a period of 3-4wk after reaching the adult stage (Patton 1978).Compared with this, in beef production about fouranimals exist in the breeding herd per market an-imal produced (Long et al. 1975, Trenkle & Will-ham 1977). The advantage in fecundity is less wheninsects are compared with poultry.

Our data have permitted a comparison of effi-ciency between an insect omnivore, A. domesticus,and conventional livestock species when both arefed the high-quality diets used in bringing the lat-ter to market condition. Despite the difficulties andlack of precise accuracy in such comparisons be-

Naracerr & DrFolrenr: Drnrs non Housr Cnrcrpr

SE

Table 5. Feed conversion efficiency (ECI) of crickets and vertebrate livestock, calculated to include losses resultingfrom carcass dressing pereentage and carcass refusea

Factors cricket Broiler chicknymph Pig

A.B.C.D.E.

Whole-body live weight ECI (%)A minus reciprocal of dressing perc-entag/B minus carcass trim (:adjusted *tit' *t ECIp ,

Whole-body dry wt ECI (%)dC minus water (:adjusted dry wt ECI)e

92 48 29.5 t8 t4.5e2 (0) 33.6 (30) re.s (34) e.7 (46) 8.6 (4r)73.6 (20) 22.8 (32) 1s.4 (2r) <9.7c 7.3 (r5)26.4 (7t.3) 17.3 (64) r4.4 (sr.s) 8.4 (ss.2) 6.8 (s2.8)2l.r (71.3) e.e (56.6y s.5 (44.7y 4.8 (50.3y 3.6 (51.3y

d Values in this table relating to whole body live weight ECI, dressing percentage,.careass trim, and (except as noted) water contentare averages of the corresponding values given in Table 3. An exception to this is the whole-body live weight ECI of 48% for thebroiler chick (see text for explanation).

b Percentage loss from values given in the preceding row shown in parentheses.c No data available.d Percentage loss from values given in Row A shown in parentheses.e Percentage loss from values given in Row C shown in parentheses.J Average moisture in cuts classed as "lean and fat." Source: Nutritive Value of Foods, 1970, Consumer and Food Economics Research

Division. Meat from which the fat has been trimmed off in the kitchen or on the plate contains more water than shown here for meatclassed as "lean and fat."

896

tween animals having widely differing dietary andhusbandry requirements, it is apparent that A. do-nxesticus is relatively efficient in its food resourceutilization compared with conventional terrestrialIivestock. The food conversion efficiency of animalsis one of the important factors that must 6e con-sidered in choosing environmentally sound foodalternatives for the future. Edible insects are amongthe alternatives (DeFoliart 1989). The ECI of in-sects varies widely depending on species and typeof food consumed (e.g., forbs, grasses, woody plantherbage, wood, organic wastes, etc.) (Slansky &Rodriguez 1987) and, considering that insects aretraditional foods in many cultures, comparisons be-tween insects and vertebrate species are needed ina variety of different resource utilization situations.

Acknowledgment

This research was supported in part. by the ResearchDivision, College of Agricultural and Life Sciences, andby the Graduate School, University of Wisconsin, Mad-ison.

References Cited

Berg, R. T. & R. M. Butterffeld. L976. New conceptsof cattle growth. Wiley, New York.

DeFoliart, G. R. f989. The human use of insects as

food and as animal feed. Bull. Entomol. Soc. Am. 35:oo a<

Dreyer, J. J. & A. S. Wehmeyer. 1982. On the nu-tritive value of mopanie worms. S. Afr. J. Sci. 78: 33-s5.

Ensminger, M. E. 1980. Poultry science, 2nd ed. In-terstate, Danville, Ill.

Finke, M. D., G. n. DeFoliart & N. J. Benevenga. 1989.Use of a four-parameter logistic model to evaluatethe quality of the protein from three insect specieswhen fed to rats. J. Nutr. ll9: 864-871.

Long, C. R., T, C. Cartwright & H. A. Fitzhugh' Jr.1975. Systems analysis of sources of genetic andenvironmental variation in efficiency of beef pro-duction: cow size and herd management. J. Anim.Sci. 40: 409-420.

Lovell, R. T. L979. Fish culture in the United States.Science 206: 1368-1372.

Meyero J. H. & A. 0. Nelson' 1963. Efficiency of feedutilization by various animal species fed similar ra-tions. J. Nutr. 80:343-349.

Nakagaki, B. J., M. L. Sunde & G. n. DeFoliart. 1987.Protein quality of the house cricket Acheta domes-tdcas, when fed to broiler chicks. Poult. Sci. 66: 1367-1371.

JounNer- oF EcoNoMIC ENToMoLocY Vol. 84, no. 3

National Research Council. 1977. Nutrient require-ments of domestic animals. No. 1: Nutrient require-ments of poultry. National Academy of Sciences,Washington, D.C.

19A2. United States-Canadian tables of feed com-- position. National Academy of Sciences, Washington,D.C.

Neville, P. F., P. C. Stone & T. D. Luckey. f961.Cricket nutrition. II. An unidentified factor in thenutrition of Acheta domesticus , J. Nutr. 7 4: 265-273.

Ozimeko L., W. C. Sauero V. Kozikowski, J. K. RyanoH. Jorgensen & P. Jelen. 1985. Nutritive valueof protein extracted from honey bees. J. Food Sci. 50:1327-1329, 1332.

Patton, R. L. 1963. Rearing the house cricket Achetadomesticus, on commercial feed. Ann. Entomol. Soc.Am. 56: 25O-25I.

196?. Oligidic diets tor Acheta domesticus (Orthop-tera: Gryllidae). Ann. Entomol. Soc. Am. 60: 1238-1242.

l9?8. Growth and development parameters for Ach-eta domesticus. Ann. Entomol. Soc. Am. 7l: 4O-42.

Roe, R. M., C. V. Clifrord & J. P. Woodring' l9BO.The effect of temperature on feeding, growth, andmetabolism during the last larval stadium of the fe-male house cricket, Acheta domesticus. J. InsectPhysiol. 26:639-644.

Slansky, F., Jr. & J. G. Rodriguez. L987. Nutritionalecology of insects, mites, spiders, and related inver-tebrates. Wiley, New York.

Taylor, R. L, & B. J. Carter' 1976. Entertaining withinsects, or: the original guide to insect cookery. Wood-bridge, Santa Barbara, Calif.

Trenkle, A. & R. L. Willham. 1977. BeeI productionefficiency. Science I98: 1009-]015.

Unired States Department of Agriculture (USDA). 197?.Comparisons among treatment means in an analysisof variance. USDA-ARS, Science Education Admin-istration, Hyattsville, Md.

Waldbauer, G. P. 1968. The consumption and utili-zation of food by insects, pp. 229-288. In J.W.L.Beament, J. E, Treherne & V. B. Wigglesworth [eds.],Advances in insect physiology, vol. 5. Academic, Lon-don.

Vhittemore, C. T., & F. V. Elsley. 1976. Practicalpig nutrition. Farming, Ipswitch, Suffolk.

Voodring, J. P., R. M. Roe & C. W. Clifiord' 1977.Relation of feeding, growth, and metabolism to agein the larval, female house cricket. J. Insect Physiol.23:2O7-2I2.

Voodring, J. P., C. W' Clifrord & B. R. Beckman.l9?9. Food utilization and metabolic efficiency inlarval and adult house crickets. J. Insect Physiol. 25:903-912.

Receioed for publicatton 76 JanuarE 7990; accepted27 Nooember 7990.